On the Importance of Label Quality for Semantic Segmentation
Name
Zlateski_On_the_Importance_CVPR_2018_paper.pdf
Description
Accepted version
Size
3.72 MB
Format
Adobe PDF
Checksum (MD5)
8b4dfc05a285971d511240970af3b2d6
Author(s)
Zlateski, Aleksandar
Jaroensri, Ronnachai
Sharma, Prafull
Durand, Frederic
Date Issued
December 2018
Journal
2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition
Publisher
IEEE
Citation
Zlateski, Aleksandar et al. "On the Importance of Label Quality for Semantic Segmentation." 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, June 18-23, 2018, Salt Lake City, UT, IEEE, December 2018. © 2018 IEEE.
Version
Author's final manuscript
Abstract
Convolutional networks (ConvNets) have become the dominant approach to semantic image segmentation. Producing accurate, pixel-level labels required for this task is a tedious and time consuming process; however, producing approximate, coarse labels could take only a fraction of the time and effort. We investigate the relationship between the quality of labels and the performance of ConvNets for semantic segmentation. We create a very large synthetic dataset with perfectly labeled street view scenes. From these perfect labels, we synthetically coarsen labels with different qualities and estimate human-hours required for producing them. We perform a series of experiments by training ConvNets with a varying number of training images and label quality. We found that the performance of ConvNets mostly depends on the time spent creating the training labels. That is, a larger coarsely-annotated dataset can yield the same performance as a smaller finely-annotated one. Furthermore, fine-tuning coarsely pre-trained ConvNets with few finely-annotated labels can yield comparable or superior performance to training it with a large amount of finely-annotated labels alone, at a fraction of the labeling cost. We demonstrate that our result is also valid for different network architectures, and various object classes in an urban scene.
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1109/cvpr.2018.00160
